What does the empty `()` do on a Matlab matrix?

Question

In Matlab, what exactly does the expression M() do where M is a matrix?

>> M = magic(3);
>> M() 

ans =

     8     1     6
     3     5     7
     4     9     2

Is the expression isequaln(M, M()) true under all circumstances? Is M() simply a copy of M, or an identical expression, or is there any context where referring to M() means something else than referring to M? Maybe in the case of operator overloading?

Answer 1

Besides the fact the it would give the default operation on some function, such as rand(), and easter eggs such as imagesc() and spy() (this will work also without the ()) , it seems to be a more efficient way to access whole arrays as long as their dimensionality is below 5 (as @Rody Oldenhuis spotted) . For example:

a=rand(2^12);

tic
for j=1:1e5
a ;
end
toc

tic
for j=1:1e5
a(:)  ;
end
toc

tic
for j=1:1e5
a()   ; 
end
toc

yield:

Elapsed time is 0.047250 seconds.
Elapsed time is 0.022260 seconds.
Elapsed time is 0.011925 seconds.

However, for assignments there's very little difference between a1=a vs a1=a(), where the latter is slower by 1.5%...

Perhaps this thread will answer some of your question regarding operator overloading.